Sustav za daljinsko otkrivanje sječe šuma pomoću zvuka

master's thesis

• Title (english): Remote deforestation detection system using sound
• Author: Ivan Filipović
• Mentor: Vladimir Čeperić (mentor)
• Committee member: Vladimir Čeperić (predsjednik povjerenstva)
• Committee member: Marin Šilić (član povjerenstva)
• Committee member: Tvrtko Mandić (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2024-07-15, Croatia

Abstract

Ovaj rad opisuje sustav za otkrivanje ilegalne sječe šume. Detekcija ilegalne sječe temelji se na zvučnim zapisima koji se s pomoću mikrokontrolera prikupljaju i obrađuju. Ako su zadovoljeni uvjeti i smatra se da je prisutna motorna pila, šalje se obavijest. Za prikupljanje zvukova koristi se mikrofon spojen na i2s sučelje, a za obradu snimki koriste se digitalni procesirajući blokovi MFE i MFCC kao i umjetna konvolucijska neuronska mreža. Slanje obavijesti šalje se putem LoRaWAN mreže korištenjem LoRa modema spojenog na SPI sučelje mikrokontrolera STM32H7A3VIT6.
Uspješno je realiziran uređaj s točnošću predikcije od 93.91 % uz potrošnju ispod 10 μAh po periodu snimanja i klasifikacije.

Abstract (english)

This paper describes a system for detecting illegal logging. The detection of illegal logging is based on audio recordings that are collected and processed using a microcontroller. If the conditions are met and a chainsaw is detected, a notification is sent. A microphone connected to the I2S interface is used to collect the sounds, and digital processing blocks MFE and MFCC, as well as an artificial convolutional neural network, are used to process the recordings. Notifications are sent via the LoRaWAN network using a LoRa modem connected to the SPI interface of the STM32H7A3VIT6 microcontroller.
A device with a prediction accuracy of 93.91% was successfully implemented, with a consumption of less than 10 μAh per recording and classification period.

• Keywords: klasifikacija zvuka; LoRa; LoRaWAN; STM32; STM32H7; konvolucijske neuronske mreže; neuronske mreže; i2s; i2c; SPI; mikrokontroler; internet stvari
• Keywords (english): audio classification; LoRa; LoRaWAN; STM32; STM32H7; Convolutional neural networks; neural networks; i2s; i2c; SPI; microcontroller; internet of things
• Language: croatian
• URN:NBN: urn:nbn:hr:168:471234
• Study programme: Title: Computing; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra inženjer/inženjerka računarstva (magistar/magistra inženjer/inženjerka računarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2024-10-12 20:34:40